Telecommunications companies are already preparing for 6G by investigating how neuromorphic artificial intelligence can curb rising energy demands. This week Ericsson and Forschungszentrum Jülich signed a Memorandum of Understanding to jointly research these brain-inspired computing architectures and their role in next-generation networks.
Future network solutions will need to deliver high intelligence and performance while minimizing energy consumption. Neuromorphic computing—hardware that emulates brain-like processes—offers a promising approach for handling complex telecom workloads more efficiently than conventional processors. Such architectures could play a critical role in enabling energy-conscious, highly responsive mobile networks.
Overcoming processing bottlenecks at the edge
The collaboration brings together Ericsson’s telecommunications domain knowledge and Jülich’s high-performance computing (HPC) expertise. The partners will study advanced computing paradigms, with particular emphasis on neuromorphic AI and HPC solutions to support the ongoing evolution of 5G and to lay the groundwork for 6G networks.
Commercial 6G services are anticipated around 2030. As artificial intelligence becomes integral to network management and operations, the joint team will evaluate AI models and techniques that can improve Ericsson’s core network and Radio Access Network (RAN) capabilities while keeping energy use in check.
A key objective is to design and prototype highly efficient computing solutions for AI inference at the radio and edge. Modern mobile systems rely on Massive MIMO, where numerous devices communicate simultaneously through many antennas. By investigating novel system architectures—including neuromorphic elements like memristors—researchers aim to accelerate optimization tasks and reduce energy consumption compared with traditional approaches.
Exploring how neuromorphic AI limits power demands in 6G networks
Nicole Dinion, Head of Architecture and Technology for Cloud Software and Services at Ericsson, highlights the strong link between AI and energy efficiency in future mobile networks. She notes that combining Jülich’s supercomputing and applied physics strengths with Ericsson’s telecom experience creates an opportunity to explore architectures that will shape next-generation communications.
Where security and commercial constraints allow, the teams may leverage JUPITER—the Jülich Supercomputing Centre’s most powerful European supercomputer—for large-scale model training. The research will also draw on Modular Supercomputing Architecture (MSA) concepts developed for exascale systems at Forschungszentrum Jülich.
Preparing infrastructure for 2030
Beyond chip-level innovation, the partnership will address operational strategies to improve energy efficiency in HPC and cloud deployments, including techniques such as heat recovery. The collaboration includes systematic benchmarking of AI approaches—especially neuromorphic methods—across Ericsson products to evaluate execution speed, scalability for large datasets, information retention, and storage efficiency.
The teams will also examine cloud deployment models inspired by concepts from the EuroHPC ecosystem, which aims to build world-class supercomputing infrastructure in Europe. These assessments will help determine feasible cloud strategies for high-performance AI workloads in telecom settings.
Professor Laurens Kuipers, a member of the Executive Board at Forschungszentrum Jülich, emphasizes the partnership’s potential to advance sustainable digital infrastructure. By combining excellence in HPC and research into neuro-inspired computing with Ericsson’s telecom expertise, the collaboration seeks to create more energy-efficient network solutions and strengthen a sovereign European digital backbone.
Evaluating vendor roadmaps for neuromorphic AI integration will give infrastructure planners an earlier understanding of how to adopt these technologies before commercial 6G rollouts commence.
See also: What telecom operators are learning from deploying AI agents at scale
Interested in AI and big data insights from industry leaders? The AI & Big Data Expo runs in Amsterdam, California, and London and is part of the broader TechEx series. It brings together panels and sessions across AI, IoT, cloud and cybersecurity to help professionals learn about emerging trends and practical deployments.
Telecoms content is produced by TechForge Media. Explore upcoming enterprise technology events and webinars hosted by the same organizers to stay informed on developments in AI, networking and related fields.